Clinical characteristics of COVID-19 patients hospitalized at Clinique Ngaliema, a public hospital in Kinshasa, in the Democratic Republic of Congo: A retrospective cohort study

Objectives To describe the clinical characteristics of patients infected with SARS-CoV-2 at Clinique Ngaliema, a public hospital, in Kinshasa, in the Democratic Republic of Congo (DRC). Methods This retrospective study analyzed medical records including socio-demographics, past medical history, clinical manifestation, comorbidities, laboratory data, treatment and disease outcome of 160 hospitalized COVID-19 patients, with confirmed result of SARS-CoV-2 viral infection. Results The median age of patients was 54 years (IQR: 38–64), and there was no significant gender difference (51% of male). The most common comorbidities were hypertension (55 [34%]), diabetes (31 [19%]) and obesity (13 [8%]). Fever (93 [58%]), cough (92 [57%]), fatigue (87 [54%]), shortness of breath (72 [45%]) and myalgia (33 [21%]) were the most common symptoms, upon admission. Patients were categorized into mild (92 [57%]), moderate (19 [12%]) and severe (49 [31%]). Severe patients were older and were more likely to have comorbidities, compared to mild ones. The majority of patients (92% [147 of 160]) patients received hydroxychloroquine or chloroquine phosphate. Regression model revealed that older age, lower SpO2, higher heart rate and elevated AST at admission were all risk factors associated with in-hospital death. The prevalence of COVID-19 and malaria co-infection was 0.63% and 70 (44%) of all patients received antimalarial treatment before hospitalization. Conclusion Our findings indicated that the epidemiological and clinical feature of COVID-19 patients in Kinshasa are broadly similar to previous reports from other settings. Older age, lower SpO2, tachycardia, and elevated AST could help to identify patients at higher risk of death at an early stage of the illness. Plasmodium spp co-infection was not common in hospitalized COVID-19 patients.


Site organization
At the start of the epidemic, each of the 25 municipalities of Kinshasa was assigned to different hospitals selected for COVID-19 patients care and designated as COVID-19 treatment center (CTC).
Clinique Ngaliema, a national secondary referral public hospital, located in the municipality of Gombe, was assigned to provide care to COVID-19 patients coming from six municipalities of Kinshasa, including: Gombe, (the municipality currently bearing the highest number of COVID-19 cases), Kintambo, Kinshasa, Lingwala, Barumbu and Ngaliema. These six municipalities represent the catchment area of the Clinique Ngaliema, as a CTC.
Clinique Ngaliema has dedicated a building with a capacity of 27 patients (with possibility to extend to 45 patients) for COVID-19 care and a staff of 107 health professionals. Suspected COVID-19 patients are transferred to the dedicated COVID-19 building, where throat and nasopharyngeal swabs are immediately collected and sent to the National Institute of Biomedical research for RT-PCR analyses.

Data collection
The following data were collected from patient medical records: demographic information, medical history, exposure history, clinical signs, symptoms and comorbidities, date of symptoms onset, as well as their severity, laboratory examination results, treatment, and disease outcome (death or discharge).
In accordance with to the National guidelines for COVID-19 management in DRC [12], the patients were categorized into mild, moderate and severe clinical groups. The mild clinical group had mild clinical symptoms including: fever, respiratory rate (RR) between 12 and 20 breaths/minute, mild respiratory tract symptoms (rhinorrhea, sore throat and cough), arthralgia, digestives symptoms (diarrhea and vomiting). The moderate clinical group presented: fever, RR between 20 and 30 breaths/ minute, oxygen saturation (SpO2) between 90 and 95%. The severe clinical group presented fever (T˚� 38.5˚C), severe respiratory distress, dyspnea, RR � 30 breaths/minute, an oxygen saturation of <90%, Acute respiratory distress syndrome (ARDS), sepsis and septic shock.

Data analysis
Data were entered and stored in Microsoft Excel 2016. Frequencies, rates and percentages were used to summarize categorical variables, whose proportions were compared using a X 2 or Fisher's exact test. Continuous variables were described using the median with interquartile range(IQR). Differences in median values were assessed using the Mann-Whitney test. Values of p less than 0.05 were considered to be statistically significant. Multivariate logistic regression models were constructed to identify factors associated with severity of symptoms, in-hospital death and those associated with unnecessary antimalarial drugs administration. Statistical analyses were done using SPSS statistical program, version 24 (SPSS, Chicago, IL, USA).

Ethical considerations
The investigators agreed to conduct the present study in full agreement with the principles of the Declaration of Helsinki' and its subsequent relevant amendments. The study was approved by the institutional ethics board of the Clinique Ngaliema. The access to patient medical records was granted by the director of the hospital. All data were fully anonymized before they have been accessed. Patients whose medical records were selected for analysis sought treatment from march to July 2020 and their data were accessed from Jun to July 2020.

Coagulation function
The level of D-dimer was above 1000 ng/mL in 107 cases (67%) and no significant difference of D-dimer concentration was found between clinical groups (Table 3).

Factors affecting in-hospital death
Of all patients included in the study, 32 (20%) died during hospitalization and 128 (80%) were discharged. In univariate analysis, odds of in-hospital death was higher in patients who did not receive hydroxychloroquine (or chloroquine phosphate) plus Azithromycin, patients with systemic inflammatory response syndrome (SIRS), higher RR, shortness of breath, elevated Ddimer and PCT or higher quick sequential organ failure assessment (qSOFA) score. Age, comorbidity, fever, SpO2, neutrophil and lymphocyte rates, CRP, AST, urea and creatinine were also associated with death ( Table 4). The regression model showed that older age, lower    SpO2, higher heart rate and elevated AST at admission were independent risk factors associated with in-hospital death (Table 4).

COVID-19 and malaria co-infection
Only one case (0.63%) of COVID-19 and malaria co-infection was observed. The parasite density was 16,900/μL (Table 3). Antimalarial treatment before hospitalization was reported by 70 (44%) cases and more than half (61%) of them were found in the moderate and severe groups ( Table 1). Factors associated with antimalarial use are presented in the appendix (see web-only S1 Table).

Discussion
This retrospective study portrayed the clinical features of COVID-19 patients hospitalized at Clinique Ngaliema and living in Kinshasa, in the DRC. The median age was found to be 54 years and severe patients were older than mild patients. Previous studies reported a broadly similar age distribution [13,14]. Male and female were equally represented. Equal sex distribution was also reported by Suxin Wan et al [15], although many studies in contrast reported gender imbalance, with either higher male to female ratio [5,16] or higher female to male ratio [4]. Interestingly, in line with other reports [17], it was noted that female were less likely to have severe symptoms compared to male. However, our results suggested that female seemed to be less affected by the infection because they attended hospital at its early stage, characterized by mild symptoms [18], compared to male who reached hospital at the progressive stage of the disease [18]. In addition, to support this observation, this study did not find any association between gender and in-hospital death.
Of all patients, 43% had a clear history of exposure to a confirmed or probable COVID-19 case. These patients were prone to visited directly from home a CTC, and at an earlier stage of the illness, compared to those with unclear history of exposure. These findings could suggest a higher level of awareness about COVID-19 existence and related symptoms in patients living in the city center of Kinshasa, who visited Clinique Ngaliema. Therefore, efforts should be invested in order to increase awareness in peri-urban areas of Kinshasa and in other provinces of DRC affected by COVID-19, where awareness on COVID-19 disease is expected to be lower.
Almost half of all patients (44%) reported having received antimalarial treatment before their hospitalization at the Clinique Ngaliema. However, it was unclear whether they have been diagnosed for malaria or they received antimalarial drugs on the basis of underlying fever or other malaria related symptoms. As malaria and COVID-19 share a couple of symptoms, a number of mild COVID-19 cases may go undetected and treated as Plasmodium spp infections. These findings underline the complexity of the of COVID-19 infection management in malaria endemic areas. Next to that, in line with other reports [14], more than half of patients received antibiotic therapy. These findings are a matter of great concern and underline the urgent need for strong guidelines in order to reduce and prevent antibiotic and antimalarial overuse during COVID-19 pandemic, in sub-Saharan Africa. Comorbidity was reported by almost half (46%) of all patients and the two most common were hypertension, and diabetes. This agrees with other studies [5]. No significant difference of comorbidity was observed between sex. In line with previous studies [19], patients with any underlying comorbidity were more likely to have severe symptoms and among reported comorbidities, hypertension was more frequently observed in severe patients. However, in contrast with other findings suggesting that comorbidities, mainly pre-existing cardiovascular disease could be associated with COVID-19 mortality [19], hypertension and heart disease were not associated with in-hospital death.
Fever, cough and fatigue were the top three common symptoms reported upon admission. These symptoms were also reported following the same order, by previous reports [13]. This list of common symptoms may help medical staff to recognise suspected COVID-19 cases, and encourage self-isolation or hospitalization, especially in remote areas with very poor COVID-19 detection capacities. Nevertheless, other non-common symptoms deserve similar attention than the more common ones.
Severe patients had significantly higher leucocyte count, elevated neutrophil rate, and lower lymphocyte rate, compared to other clinical groups. This has been also documented elsewhere [20]. Our study also revealed that most patients had elevated PCT and CRP, and higher CRP concentration was also associated with the disease severity. These inflammatory abnormalities were also reported previously [21].
Coagulation disorders, measured by the level of D-dimer were observed in more than half patients (67%). Elevated D-dimer was also reported in previous studies [22]. In line with other previous findings [23], our study also revealed that elevated D-dimer was associated with the disease severity.
Liver enzymes abnormalities were found in 66% of patients. This higher proportion of liver dysfunction on admission could be explained, by the observed time interval (�7 days) from initial symptoms to hospitalization, which coincides with the time of occurrence of severe symptoms [13]. High proportions of AST and ALT elevation were also reported in other studies [24]. Previous studies have also shown that abnormalities of liver function, especially elevation of AST were significantly associated with COVID-19 severity [25] and mortality [26]. Our study also documented that severe group had the highest concentration of AST.
We observed a hospital case fatality rate of 20%, which seemed to be lower than those reported by other studies [14]. However, this is unlikely to reflect the true fatality rate of the disease, as out-patients and those with missing data were excluded. Independent risk factors associated with in-hospital death in this study included older age, lower SpO2, higher heart rate and higher AST level. Association of age and death [14] as well as lower SpO2 and death was also reported in China [27]. Oxygen saturation has appeared to be a reliable surrogate marker of COVID-19 illness severity and death. However, to the best of our knowledge, heart rate was only described as a predictor for a positive SARS-CoV-2 test [17].
This study reported a low prevalence (0.63%) of malaria and COVID-19. A couple of reasons may support this observation. First, the lower prevalence of malaria observed in the city center (the catchment area of Clinique Ngaliema and its neighboring communities) compared to peri-urban areas [28]; and second, the considerable proportion of patients who received antimalarial drugs before hospitalization. This is to our knowledge the first study reporting the coexistence of malaria and COVID-19 infections. Nevertheless, previous studies also have shown that co-infections including other viral, bacterial and fungal infections were rarely found among patients with COVID-19 [29].
The study has a number of limitations. First, lack of some key data (laboratory tests) and excluded variables from analysis because of missing values (high sensitive troponin, ferritin, blood gas, ionogramm etc.) that could play an important role, due to the retrospective design. Second, excluded cases, could lead to a selection bias. However, this is unlikely given that patients excluded were similar to those included in the study, in regard to all socio-demographic data. Last, although nearly half of patients included were located out of the catchment area of Clinique Ngaliema, our sample may not be representative of Kinshasa. Therefore, our findings should be generalized to COVID patients of Kinshasa with caution.
Supporting information S1